Project Summary/Abstract Patients with persistent post-concussion vestibular dysfunction (PCVD) demonstrate visual motion sensitivity characterized by visual and motion stimuli inducing vestibular symptoms. We hypothesize that patients may acutely benefit from altered weighting of multisensory, particularly visual, input into vestibular processing networks to compensate for central or peripheral vestibular impairment, although this may persist and become maladaptive, leading to persistent vestibular symptoms. However, the changes in multisensory processing that underlie PCVD are largely theoretical and represent a significant knowledge gap in our understanding. In a recent pilot study, we found selective increased activation in the primary vestibular cortex and vestibular multisensory processing regions in patients with subacute PCVD during a novel task-based fMRI visual- vestibular paradigm as well as altered resting-state fMRI connectivity between visual and vestibular processing centers, which correlates with symptom severity. Our central hypothesis is that persistent PCVD is due to altered multisensory vestibular processing, with increased activation/connectivity of visual and oculomotor inputs into the vestibular network. Prior efforts to classify concussion based upon clinical symptoms have been limited as pre-existing symptoms may mimic post-concussion symptoms. We hypothesize that defining endophenotypes for PCVD patients using a combination of clinical and neuroimaging metrics will better subdivide this population and will correlate with response to vestibular rehabilitation therapy (VRT). To assess these hypotheses, we propose the following three Specific Aims: (1) define regional brain activation that distinguishes PCVD patients from concussion recovered and control groups using a novel visual- vestibular task-based fMRI paradigm and correlations with subjective and objective vestibular testing; (2) identify alterations in functional networks and dynamic states in the PCVD group compared to recovered and control groups at rest and correlations with subjective and objective vestibular testing; and (3) stratify PCVD subjects into endophenotypes using clinical and neuroimaging metrics and determine the predictive power of these endophenotypes to predict VRT response. We propose a longitudinal study with three groups: (1) subacute PCVD patients, (2) patients with prior symptomatic concussion but who have since recovered, and (3) healthy controls. All subjects will undergo comprehensive vestibular testing and brain MRI. PCVD subjects will repeat clinical testing after completion of VRT. We expect to confirm and expand our preliminary data findings as well as develop endophenotypes that will be predictive model of VRT response. The proposed study will fill in current gaps in knowledge, drive the development of novel therapies, identify neuroimaging/clinical patterns that predict therapy response, and lead to the development of tailored, ...